OF THE GENUS SYMPHYTUM. 499 
The next attempt to group the species was made by Otto Kunze (PI. 
orient.-rossic. in Act. Hort. Petropol. x. 1887, p. 219). He reduces the 
number of species to five, and apparently regards all the others as 
synonymous with these. His “ species " are :—— 
1. S. bulbosum including S. ottomanum. 
2. S. tuberosum —,, S. sylvaticum. 
3. S. tauricum » S. grandiflorum. 
4. S. officinale » S. uliginosum, caucasicum and mediterraneum. 
5. S. orientale „ S. peregrinum, asperum, palestinum, kurdicum, 
anatolicum, brachycalyx, sepulerale and grandiflorum, hort. nee DC. 
I altogether fail to see how any characters could justify such an 
incongruous grouping of the species as is here displayed. 
The arrangement by Gürke in Engler & Prantl’s * Pflanzenfamilien? 
follows that of Boissier, but adds S. cordatum, Waldst. & Kit., to the 
group of S. officinale. 
Finally, Kusnezow in the * Caucasian Species of the Genus Symphytum? 
(1910), retains Boissier's main divisions, but carries the subdivision of these 
a step further. The following is an outline of his system :— 
I. EUSYMPHYTUM. 
i. Cyanea. А. S. officinale, uliginosum, mediterraneum, | molle, cau- 
casicum. B. S. orientale, brachycalyx, palestinum, | kurdicum, 
anatolicum, sepulerale, asperum, peregrinum. 
ii. Ochroleuca. А. S. tuberosum, Gussonei, B. S. cordatum, grandi- 
Jlorum, tauricum, sylvaticum. 
П. Всівоѕсм. S. bulbosum, ottomanum. 
The weak point in all these arrangements is that the main sections are 
founded on a comparatively unimportant character, viz., the relative length 
of the corolla-scales and the corolla. From this it results that S. bulbosum 
and 8. ottomanum, in which the scales are longer than the corolla, are 
associated together, and each is separated from the species to which it is 
naturally allied. To the characters of the groups adopted by Kusnezow the 
same objection applies, with the further disadvantage that they are not 
sufficiently definite and admit of numerous exceptions, especially with regard 
to the branching of the stem and the colour of the flowers. 
I now submit the system of classification which a study of actual specimens 
of all the known species (except the doubtful S. sieyosmum), many of them 
in the living state, has led me to adopt, and which has already been outlined 
when treating of the charaeters used in the determination of the species. 
